home edit page issue tracker

This page pertains to UD version 2.

It appears that you have Javascript disabled. Please consider enabling Javascript for this page to see the visualizations.

Metadata harmonisation: align spoken-data fields with UniDive naming conventions

Back to Zazaki ZSD · Back to index

Repo: https://github.com/UniversalDependencies/UD_Zazaki-ZSD

Cross-posting from the UniDive WG1 T1.5 (spoken language guidelines) metadata harmonisation review. We compared UD_Zazaki-ZSD’s current CoNLL-U metadata against the proposed naming conventions (see also the full treebank status table). This is a suggestion for maintainers to review - feel free to push back on anything that doesn’t fit the corpus. The comparison was carried out semi-automatically with the help of Claude (Anthropic); errors or misunderstandings are possible, so please double-check anything unclear.

1. Document-level (naming conventions)

No document_id exists, but it’s trivial to derive: sent_id follows Seyristane_dialogue_<number><A/B> (e.g. Seyristane_dialogue_171A), and the whole corpus (200 sentences) is a single interview/dialogue.

Field	Suggestion
—	add `# document_id = Seyristane_dialogue` corpus-wide (single document)

2. Sentence-level (naming conventions)

Field	Suggestion
`text_en`	rename to `text_eng` (ISO 639-3)

Implementation notes

Quick search & replace: text_en → text_eng: python3 workgroups/spoken-data/scripts/harmonize_metadata.py rename-comment UD_Zazaki-ZSD --map text_en=text_eng --write.
Needs a small script: add # document_id = Seyristane_dialogue corpus-wide.
```
python3 workgroups/spoken-data/scripts/harmonize_metadata.py derive-document-id \
    UD_Zazaki-ZSD --pattern '^(?P<doc>Seyristane_dialogue)_\d+[AB](?:_split\d+)?$' --write
```
Verified against the local clone (dry-run, all three splits): the simpler pattern from the issue draft (^(?P<doc>.+)-\d+$-style, without the _split\d+ suffix) misses 2 sent_ids in the test split (Seyristane_dialogue_185A_split1/_split2) - the pattern above (with the optional _split\d+ suffix) correctly derives the single document across all three files with no unmatched sent_ids.

This issue was prepared as part of the UniDive WG1 T1.5 spoken language guidelines effort. Happy to help implement these changes ourselves if that’s easier than doing it on your end - just let us know.